- Email: [email protected]
Microstructure and creep behavior of FGH95 nickel-base superalloy
Materials Science and Engineering A 528 (2011) 2076–2084
Contents lists available at ScienceDirect
Materials Science and Engineering A journal homepage: www.elsevier.com/locate/msea
Microstructure and creep behavior of FGH95 nickel-base superalloy Tian Sugui a,∗ , Xie Jun a , Zhou Xiaoming b , Qian Benjiang a , Lun Jianwei a , Yu Lili a , Wang Wuxiang b a b
School of Materials Science and Engineering, Shenyang University of Technology, Shenyang 110870, China Beijing Institute of Aeronautical Materials, Beijing 100095, China
a r t i c l e
i n f o
Article history: Received 30 July 2010 Received in revised form 10 November 2010 Accepted 10 November 2010
Keywords: FGH95 nickel-base superalloy Microstructure Creep Deformation mechanism Fracture feature
a b s t r a c t By means of the measurement of creep curves and the microstructure observations, the influence of the solution temperatures on the creep behavior of FGH95 nickel-base superalloy is investigated. Results show that, after solution treated at 1150 ◦ C, some coarser ␥ precipitates are distributed in wider boundary regions where no fine ␥ phase is precipitated. As the solution temperature is raised to 1165 ◦ C, the grain size of the alloy increases obviously, and the carbide is continuously precipitated to form the film along the boundaries. When the alloy is solution treated at 1160 ◦ C, the coarser ␥ phase in the alloy is fully dissolved, the fine ␥ phase with higher volume fraction is dispersedly distributed within the grains, and some particles of (Nb, Ti)C are precipitated along the grain boundaries, which can effectively hinder the grain boundary slipping and dislocation moving. Thereby, the alloy displays a better creep resistance under the applied stress of 1034 MPa at 650 ◦ C. The deformation mechanism of the alloy during creep is twinning, dislocation shearing or bypassing the ␥ phase, and the 1 1 0 super-dislocation which shears into the ␥ phase may be decomposed to form the configuration of (1/3)1 1 2 super-Shockleys partial and stacking fault. In the later stage of creep, the deformation features of the alloy are the single and double orientations slipping of dislocations activated in the alloy. As the creep goes on, some dislocations piled up in the regions near the boundaries may bring the stress concentration to promote the initiation and propagation of micro-cracks, which is thought to be the fracture mechanism of the alloy during creep. © 2010 Published by Elsevier B.V.
1. Introduction With the development of aerospace and ground transportation industry, the properties of aerospace turbines are increasingly required to be improved, and especially, the turbine disks of aeroengines are required to be of high temperature tolerance and creep resistance [1]. Traditional wrought superalloys can hardly meet the requirements of aerospace turbine disks for their poor temperature tolerance and loading capacity resulted from their serious composition segregation in ingots and poor hot processability [2–5], especially the weaker cohesive force of grain boundaries [6–8]. FGH95 alloy is a nickel-base superalloy with higher alloying degree and volume fraction of ␥ -phase [6,9]. Compared with wrought superalloys, FGH95 superalloy has the capabilities of higher temperature tolerance and loading tolerance, due to their excellent characteristics, such as uniform chemical composition and fine grain size. Therefore, the alloy is an excellent material used for preparing the advanced aerospace turbine disks with high thrust-weight ratio [10,11]. The microstructure of FGH95 superalloy consists of ␥ matrix, ␥ and carbide phases. Various size, morphology and distribu-
∗ Corresponding author. Tel.: +86 24 25494089; fax: +86 24 25496768. E-mail address: [email protected] (T. Sugui). 0921-5093/$ – see front matter © 2010 Published by Elsevier B.V. doi:10.1016/j.msea.2010.11.038
tion of ␥ phase can be obtained in the alloy by different heat treatment regimes [12,13]. The deformation mechanism of the polycrystalline Ni-base superalloys during creep includes twinning, dislocations by-passing or shearing into the ␥ phase [14–16]. Actually, the mechanical properties and creep behaviors of the alloy are related to the quantities, morphology and distribution of ␥ -phase, and especially, the configuration of the boundary and carbides have an important effect on the creep resistance of the alloy [17]. For example, after the alloy is solution treated for cooling in molten salt, the total number and size of secondary ␥ phase increase, which can effectively improve the plasticity of the alloy at high temperature [18,19]. Because the various microstructures in the alloy may be obtained by different heat treatment regimes, it is very important to understand the influence of heat treatment regimes on the microstructure and creep resistance of the alloy. Although some literatures report the creep behaviors and deformation features of the powder superalloys, the influence of heat treatment regimes on the configuration and distribution of ␥ -phase and carbide in FGH95 superalloy is still not clear. In this paper, the alloy is solution treated at different temperatures, and then the creep properties are measured and the microstructure is observed by using SEM and TEM for investigating the influences of the solution temperatures on the boundary morphologies and distribution of the ␥ and carbide phases. Addi-
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084 Table 1 Heat treatment regime of FGH95 powder superalloy. Solution treated
solution temperature on the microstructure and the creep feature of the alloy.
Cooled in molten salt
Aging
583 ◦ C for 15 min
870 ◦ C × 1 h + 650 ◦ C × 24 h
◦
3. Experimental results and analysis
1150 C for 1 h 1160 ◦ C for 1 h 1165 ◦ C for 1 h
3.1. Influence of solution temperature on the microstructure
Table 2 Chemical composition of FGH95 powder alloy (mass fraction, %). C
B
2077
Cr
0.060 0.012 12.98
Co
Al
Ti
W
Mo
Nb
Ni
8.00
3.48
2.55
3.40
3.40
3.50
Bal.
tionally, the deformation mechanism and fracture feature of the alloy during creep are briefly discussed. 2. Experimental procedure FGH95 powder particles of the nickel-base superalloy with the size about 150 mesh were put into a stainless steel can for pretreating at 1050 ◦ C for 4 h. The can containing FGH95 alloy powders was hot isostatic pressing treated for 4 h under the conditions of 1120 ◦ C and 120 MPa. The various solution temperatures are selected for investigating the influence of solution temperatures on the microstructure and creep properties of the alloy, and the cooled rate of the specimen in molten salt is measured to be about 110 ◦ C/min. The selected heat treatment regimes of the alloy are listed in Table 1, and the error ranges of the used heating furnace are ±2 ◦ C. The chemical composition of FGH95 alloy is shown in Table 2. By means of the anode selective dissolving method, the volume fraction of ␥ -phase in FGH95 alloy is measured to be about 47%. After heat treated at different conditions, the ingot of FGH95 alloy was cut into the specimens with the cross-section of 4.5 mm × 2.5 mm and the gauge length of 20 mm. Uniaxial constant load tensile testing was performed, in a GWT504-model creep testing machine, for measuring creep curves under the experimental conditions of 1020–1050 MPa and 650–670 ◦ C. The yield strength of FGH95 alloy is measured to be 1110 MPa at 650 ◦ C. The strain data of the alloy at different conditions were measured with an extensometer for portraying the creep curves, twice of each creep testing were conducted to ensure the statistical confidence. The microstructures of FGH95 alloy at different states were observed under SEM and TEM for investigating the influence of the
After the alloy was solution treated at 1150 ◦ C and twice aged, the microstructure of the alloy consists of the ␥ and ␥ phases, and the average grain size of the alloy is about 10–20 m, as shown in Fig. 1(a), indicating that some coarser ␥ -particles are precipitated in the wider boundary regions, and the average size of the coarser ␥ phase is about 1–2.5 m. The magnified morphology of the alloy is shown in Fig. 1(b), indicating that significant amount of the fine ␥ particles is dispersedly distributed within the grains, the size of the ones was about 0.1–0.3 m. And no fine ␥ -particles were precipitated in the regions near the coarser ␥ -phase, the regions were defined as the depleted zone of the fine ␥ -phase as marked by the arrow in Fig. 1(b), the magnified morphology of the depleted zone of ␥ -phase is marked by the arrow in Fig. 1(c). As the solution temperature raises to 1160 ◦ C, the grain boundaries appear obviously, and the average size of the grains in the alloy is about 15–25 m as shown in Fig. 2(a). Compared to Fig. 1(a), the coarser ␥ -precipitates in the boundary regions disappear, and the white particles with size about 0.2 m are precipitated within the grains and along the boundaries as marked by the arrow in Fig. 2(a). The magnified morphology of the alloy is shown in Fig. 2(b), indicating that the coarser ␥ particles and depleted zone of the fine ␥ phase have disappeared, and the secondary ␥ phase is dispersedly distributed within the grains. The grain boundary in the alloy is marked by longer arrow and some particles are homogeneously precipitated along the boundaries and within the grains as marked by the short arrows in Fig. 2(b). As the solution temperature increases to 1165 ◦ C, the grain sizes in the alloy increase, the linear-like boundaries appear in the alloy, and the films of the white phase are continuously precipitated along the boundaries as marked by the arrows in Fig. 3(a). The magnified morphology of the white phase is shown in Fig. 3(b), indicating that white particles are continuously precipitated to form the films along the boundaries as marked by the arrow in Fig. 3(b), and significant amount of the fine secondary ␥ phase is precipitated within the grains, no depleted zones of the fine ␥ phase are detected in the alloy. The grain sizes after the alloy was solution treated at various temperatures are measured as listed in Table 3. Moreover, by means of composition analysis under SEM/EDS, it is indicated that
Fig. 1. Microstructure after the alloy solution treated at 1150 ◦ C. (a) After solution treated at 1150 ◦ C, some primary ␥ phase precipitated in the wider boundary regions, (b) fine ␥ phase distributed dispersedly within the grains, and the depleted zones of the fine ␥ phase marked by arrow, and (c) the magnified morphology of the depleted zone of ␥ -phase.
2078
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084
Fig. 2. Microstructure after the alloy solution treated at 1160 ◦ C. (a) After solution treated at 1160 ◦ C, the coarser ␥ phase disappearing and the grain boundaries appearing obviously in the alloy and (b) the secondary ␥ phase distributed dispersedly within the grain, and some particles precipitated in the alloy as marked by the arrows.
Fig. 3. Microstructure after the alloy solution treated at 1165 ◦ C. (a) After solution treated at 1165 ◦ C, the average grain size of alloy increasing obviously, and films of white phase precipitated along the boundaries as marked by arrows and (b) the secondary ␥ phase distributed dispersedly within the grains, and films of the white phase distributed along the grain boundaries. Table 3 Grain sizes of the alloy solution treated at different temperatures. Solution temp. (◦ C) Average grain sizes (m)
1150 10–20
1160 15–25
1165 20–40
the elements Nb, Ti and C are richer included in the white particles which are located within the grain and boundary regions as shown in Figs. 2 and 3, respectively. After the alloy was solution treated at 1160 ◦ C, the fine ␥ particles with the size of about 0.1 m are dispersedly precipitated within the grains as shown in Fig. 4(a), the particles can effectively hinder the dislocation movement to enhance the creep resistance of the alloy. The smaller space between the fine ␥ -particle is mea-
sured to be about 0.03 m, the bigger space between the fine ␥ -particles is measured to be about 0.12 m as marked by letters L1 and L2 in Fig. 4(a), respectively. It is indicated by TEM/EDS analysis that the elements Nb, Ti and C are richer included in the carbide particle which are located in the boundary as shown in Fig. 4(b), and the particle is identified as (Nb, Ti)C phase by means of the diffraction spots analysis as marked in Fig. 4(c). 3.2. Influence of solution temperature on creep behaviors Under the applied stress of 1034 MPa at 650 ◦ C, creep curves of the alloy after solution treated at different temperatures are shown in Fig. 5. The creep curve of the alloy after solution treated at 1150 ◦ C
Fig. 4. Morphologies of ␥ and carbide phases. (a) Fine ␥ phase precipitated dispersedly within the grain, (b) carbide particles precipitated along the boundaries, and (c) SAD patterns.
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084
where ε˙ ss is the strain rate during steady state creep, A is a constant related to the microstructure, A is the applied stress, n is the apparent stress exponent, R is the gas constant, T is the absolute temperature, and Qa is the apparent creep activation energy. After solution treated at various temperatures, the creep curves of the alloy in the ranges of 650–670 ◦ C and 1020–1050 MPa were measured. The dependence of the strain rates of the alloy during steady-state creep on the applied temperatures and stresses is shown in Fig. 6. Fig. 6(a) shows the relationship between the strain rates and the temperatures under the applied stress of 1034 MPa, and the dependence of the strain rates on the applied stresses at 650 ◦ C is shown in Fig. 6(b). According to the data during the steady state creep, the creep activation energies and stress exponents of FGH95 alloy, which is solution treated at 1150 ◦ C and 1160 ◦ C, are calculated to be QA = 510.1 ± 20 kJ/mol, QB = 580.3 ± 20 kJ/mol and nA = 15.4, nB = 14.1, respectively. After solution treated at different temperatures, the creep lifetimes of the alloy under the applied various stresses and temperatures are measured as listed in Table 4. It may be understood from Table 4 that, under the applied stress of 1034 MPa at 650 ◦ C, the lifetime of the alloy solution treated at 1150 ◦ C is measured to be 67 h. As the solution temperature enhances to 1160 ◦ C, the lifetime of the alloy increases to 104 h. When the solution temperature raises further to 1165 ◦ C, the lifetime of the alloy decreases rapidly to 9 h, as shown in Table 4. This indicates that the solution temperature has an obvious influence on the creep lifetimes of the alloy. After solution treated at 1150 ◦ C, the lifetime of the alloy is about 95 h under the applied stress of 1020 MPa at 650 ◦ C, and the strain rate during steady state creep is measured to be about 0.00827%/h. The lifetime of the alloy decrease to 67 h as the applied stress increase to 1034 MPa at 650 ◦ C, and the creep lifetime of the alloy decreases to 37 h as the applied stress increases to 1050 MPa at 650 ◦ C. And the alloy solution treated at 1160 ◦ C displays also the similar regularity under the experimental conditions. This indicates that the alloy has an obvious sensitivity on the applied stresses and temperatures.
4.0
Strain, ε (%)
o
A - solution at 1150 C o 3.2 B - solution at 1160 C o C - solution at 1165 C o 2.4 T -- 650 C σ -- 1034 MPa 1.6 0.8
A
C
0.0 0
20
40
B
60
80
100
Time, (h) Fig. 5. Under the applied stress of 1034 MPa at 650 ◦ C, creep curves of the alloy solution treated at different temperatures.
is marked by letter A in Fig. 5, through which the strain rate of the alloy during steady state creep is measured to be 0.0102%/h, the lasting time is about 40 h, and the creep lifetime of the alloy is measured to be 67 h. The creep curve after the alloy solution treated at 1160 ◦ C is marked by letter B in Fig. 5, indicating that the alloy displays a lower strain rate during steady-state creep, and the creep lifetime of the alloy is measured to be 104 h. When the solution temperature increases to 1165 ◦ C, the creep lifetime of the alloy is measured to be about 9 h, as marked by letter C in Fig. 5. Transient strain of the alloy occurs when the loading is applied at high temperature, and the strain rate of the alloy decreases as the creep goes on. The strain rate keeps constant once the creep enters the steady state stage, therefore, the strain rate of the alloy during steady state creep may be expressed by Dorn’s law as follows:
Q a
ε˙ ss = AAn exp −
(1)
RT -3.5
a
σ − 1034 ΜPa
-4.0
-4.0
QA = 510.1+ 20 kJ/mol ln( εss)
ln( εss)
-5.5
QB = 580.3 + 20 kJ/mol
-6.0
1.074
1.080
1.086
o
T -- 650 C nA = 15.4
-5.0 -5.5
nB = 14.1
-6.0
-6.5 1.068
b
-4.5
-4.5 -5.0
2079
1.092
-6.5
1.098
1/T( 10-3K-1)
6.928
6.936
6.944
ln ( σ)
6.952
6.960
Fig. 6. Dependence of the strain rate during steady state creep on the applied temperatures and stresses for the alloy solution treated by different temperatures. (a) Strain rate vs. temperatures at 1034 MPa and (b) strain rate vs. the applied stress at 650 ◦ C.
Table 4 Effect of solution temperature on stress rupture properties of FGH95 alloy. Solution temp. (◦ C)
650 ◦ C
1034 MPa
1020 MPa
1150 1160 1165
1034 MPa
660 ◦ C
1050 MPa
670 ◦ C
tf (h)
ı (%)
ε˙ (%/h)
tf (h)
ı (%)
ε˙ (%/h)
tf (h)
ı (%)
ε˙ (%/h)
tf (h)
ı (%)
ε˙ (%/h)
tf (h)
ı (%)
ε˙ (%/h)
95 156 –
4.2 3.7 –
0.00827 0.00307 –
67 104 9
3.4 2.8 1.3
0.0102 0.00367 0.00304
37 51 –
3.2 2.1 –
0.0129 0.00456 –
26 40 –
3.6 2.2 –
0.0208 0.00825 –
12 17 –
2.3 1.7 –
0.0418 0.0179 –
2080
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084
Fig. 7. Microstructure after the alloy (solution treated at 1160 ◦ C) crept for 104 h up to fracture at 650 ◦ C/1034 MPa. (a) Dislocation loops in the alloy, (b) carbide particles precipitated uniformly within the grain, (c) morphology of the super-dislocation shearing into ␥ phase marked by black arrow, the stacking fault in the region A and dislocation tangles piled up in the B region near the boundary, (d) carbide particles precipitated along the boundary, and dislocation tangles piled up in the regions near the boundaries.
3.3. Deformation features of alloy during creep After solution treated at 1160 ◦ C, the morphology of the alloy crept for 104 h up to rupture under the applied stress of 1034 MPa at 650 ◦ C is shown in Fig. 7. Fig. 7(a) displays the feature of the stacking fault as marked by white arrow, and two (1/3)1 1 2 superShockleys partials are located on the two sides of the stacking fault [15]. Some dislocation loops with various sizes appear clearly in the ␥ matrix of the alloy, the smaller dislocation loop is marked with black arrow in Fig. 7(a), and larger dislocation loop is located in the region above the stacking fault. It can be deduced by analysis that the 1 1 0 super-dislocation which shears into the ␥ or ␥ -phase may be decomposed to form the configuration of two (1/3)1 1 2 super-Shockleys partials and the stacking fault [20]. Some blocky carbides are precipitated in local areas of the alloy as marked by black arrow in Fig. 7(b). The grain boundary in the alloy is marked by white arrow in Fig. 7(c), and dislocation tangles are piled up in the region B near the boundary, which suggests that the boundaries may effectively hinder the dislocation movement during creep. Besides, the morphology of 1 1 0 super-dislocation shearing into the ␥ -phase is marked by the black arrow, and the stacking fault formed from the dislocation decomposition is marked by the letter A in Fig. 7(c). The blocky carbides are precipitated along the boundary as marked by the black arrow in Fig. 7(d),
and some dislocations are piled up in the regions near the carbides, which indicates that the carbide particles can effectively hinder the dislocation movement. But, the fact that significant amount of dislocations is piled up in the regions near the boundaries can cause the stress concentration to promote the initiation and propagation of the microcrack along the boundaries as the creep goes on. After the alloy was crept up to fracture under the applied stress of 1034 MPa at 650 ◦ C, the feature of the twinning deformation is shown in Fig. 8(a), and the twinning plane is identified as (1 1 1) plane by means of the diffraction pattern analysis as shown in Fig. 8(b). The blocky carbide is precipitated along the grain boundary as marked by shorter arrow in Fig. 9, and some micro-twinning is detected in the alloy as marked by black arrow. In another local area, the morphology and diffraction spots of the micro-twinning are shown in Fig. 9(b) and (c), this is well agreement with the result in literature [14]. The one end of the micro-twinning is stopped at the grain boundary, which suggests that the grain boundaries have an obvious effect on hindering the twinning deformation. 3.4. Fracture feature of alloy during creep After solution treated at 1160 ◦ C, the morphology of the alloy crept for different times under the applied stress of 1034 MPa at 650 ◦ C is shown in Fig. 10.
Fig. 8. Twinning morphology in FGH95 alloy during creep at 650 ◦ C/1034 MPa. (a) Twinning and (b) SAD patterns.
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084
2081
Fig. 9. Morphologies of the carbide precipitated along the boundary and micro-twinning formed in the alloy during creep under the applied stress of 1034 MPa at 650 ◦ C. (a) Carbide precipitated along the boundary, (b) micro-twinning, and (c) SAD patterns.
Fig. 10. Morphologies of slipping traces on the surface of the specimen (solution treated at 1160 ◦ C) crept for different times. (a) After crept for 60 h, a few slipping traces appearing within the different grains and (b) crept for 80 h, significant amount of the slipping traces appearing on the surface of the sample.
After the alloy is crept for 60 h, the morphology of the slipping traces on the sample surface is shown in Fig. 10(a), which displays the feature of the single orientation slipping within the grain. But the slipping traces with different directions appear within different grains, and the intersected slipping traces appear in the boundary region as marked by the arrow in Fig. 10(a), which indicates that the boundary may change the direction of the slipping traces. When crept for 80 h, the amount of the slipping trace on the sample surfaces increases obviously, as shown in Fig. 10(b), and some white particles are uniformly precipitated within the grains.
Significant amount of dislocations is piled up in the regions near the boundary and carbides during creep, as shown in Fig. 7(d), which suggests that the carbides and boundaries can effectively hinder the dislocation movement. The dislocations piled up in the regions near the boundaries may bring about the stress concentration, as the creep goes on, to promote the initiation of the micro-cracks along the boundary when the value of the stress concentration exceeds the yield strength of the boundary, as marked by the arrow in Fig. 11(a). The number of the slipping traces in the sample surface increases as the creep goes on, as shown in Fig. 11(b), and the
Fig. 11. Morphologies of the slipping traces on the surface of alloy (solution treated at 1160 ◦ C) crept up to fracture. (a) Micro-crack initiated along the boundary as marked by the arrow, (b) slipping traces in the surface, and (c) feature of cracks propagated along the boundaries.
2082
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084
Fig. 12. After solution treated at 1165 ◦ C, morphology of the alloy crept for 9 h up to fracture. (a) Crack initiated along the boundary and (b) carbide film reserved between the cracks.
Fig. 13. Schematic diagram of dislocation bowing out in the congregated regions of the ␥ particles.
slipping traces with the double-orientation feature appear within the grains, the slipping steps are displayed in the region near the boundaries as marked with the white arrow, and some cracks appear in the region near the boundary as marked with the shorter arrow in Fig. 11(b). Furthermore, when significant amount of dislocations move to the region near the boundaries again to produce the stress concentration, this may promote the propagation of the crack along the boundary, as shown in Fig. 11(c), which indicates that the fracture originated from the cracks initiating and propagating along the boundaries displays a non-smooth surface due to the pinning effect of the carbide particles precipitated along the boundaries. Therefore, it may be deduced that the carbide particles precipitated along the boundaries may restrain the cracks propagation along the boundaries for enhancing the creep resistance of the alloy. After solution treated at 1165 ◦ C and twice aged, the surface morphology of the alloy crept for 9 h up to rupture under the applied stress of 1034 MPa at 650 ◦ C is shown in Fig. 12, which illustrates that only a few slipping trace appears on the surface of the sample, and some micro-cracks are initiated along the boundaries vertical to the applied stress axis as marked by arrow in Fig. 12(a). After the sample is polished and eroded, the morphology of the crack propagated along the boundaries is shown in Fig. 12(b). It may be understood from Fig. 12(b) that the fracture after the crack is propagated displays the smooth surface, and the carbide film is reserved within the crack between the grains as marked by arrow in Fig. 12(b), which suggests that once the carbide is continuously precipitated to form the film along the boundary, which may weaken the cohesive strength between the grains to damage the creep lifetimes of the alloy. Moreover, it is identified by means of composition analysis under SEM/EDS that the elements Nb, Ti, C and O are richer included in the white particles on the surface of the samples, as shown in Figs. 10, 11 and 12, respectively, therefore, it is thought that the white particles on the surface of the samples are the oxides of the elements Nb, Ti and C.
4. Analysis on deformation features of the alloy during creep Significant amount of the fine ␥ particles is precipitated within the grains, which may effectively hinder the dislocation movement. When the deformed dislocations move over the ␥ phase during creep, the dislocation loops are kept around the ␥ particles as shown in Fig. 7(a), which suggests that the deformation feature of the alloy during creep is the dislocations moving over the ␥ -phase by Orowan bypasses mechanism. It is reasonable consideration that the various spaces between the ␥ particles appear in different regions, the dislocations may bow out along the wider channels between the two ␥ particles during creep, and the bowing dislocations move over the ␥ -phase by Orowan mechanism to encounter for forming the dislocation loops, as marked with black arrow in Fig. 7(a). When the dislocations bow out along the channels during creep, the applied stress which is enough to overcome the Orowan resistance can be expressed as follows: or =
·b L
(2)
where is the shear modulus, b is the Burgers vector, and L is the space between two ␥ particles. This indicates that the resistance of the dislocation movement increases with the diminishing of the space between the ␥ -particles, and the resistance of alloy enhances with the volume fraction of ␥ -particles. The space between the ␥ particles diminishes when some ␥ particles are congregated together, the smaller space is measured to be L1 = 0.03 m as shown in Fig. 4(a), but the bigger space between the ␥ -particles is measured to be L2 = 0.12 m. Compared to the bigger space, the dislocations bowing out in the regions with smaller space needs fourfold Orowan resistance according to the formula (2), therefore, it is difficult for the dislocations to bow out along the channels with smaller space, but they can bow out along the channels with larger space to form the larger loops as shown in
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084
Fig. 7(a). When the dislocations bow out along the channels with bigger space (L), the formation process of the bigger dislocation loops is schematically shown in Fig. 13. In the later stage of creep, significant amount of dislocations move to the region near the ␥ particles to generate stress concentration, and when the stress value originated from the stress concentration exceeds the yield strength of ␥ phase, the 1 1 0 super-dislocation may shear into the ␥ -phase, as marked by black arrow in Fig. 7(c). Furthermore, the 1 1 0 super-dislocations may be decomposed to form the configuration of two (1/3)1 1 2 superShockleys partials and the stacking fault, as marked by letter A in Fig. 7(c). The critical stress of dislocation shearing into ␥ phase increases with the yield strength. And the critical stress () can be expressed as follows [22]: cs =
APB b
0.3 ·
APB
T
·f ·r
2083
Because the boundaries and the carbide particles can effectively block the dislocation movement, and especially, the carbides may restrain the boundaries slipping during creep, it may be concluded that the carbide particles precipitated along the boundaries have an important effect on improving the creep resistance of the alloy. Although the carbide particles precipitated along the boundaries can improve the cohesive strength of the boundaries, the microcracks are still initiated and propagated along the boundaries, which suggests that the boundaries are still the weaker regions for causing fracture of the alloy during creep. And once, the carbide is continuously precipitated to form the film along the boundary, which may weaken the cohesive strength between the grains to damage the creep lifetimes of the alloy. The analysis is in agreement with the experimental results stated above.
1/2 (3)
where T is the dislocation line tension, r is the radius of the ␥ particle, b is the Burgers vector, f is the volume fraction of ␥ phase, and APB is the antiphase boundary energy originated from the dislocation shearing ␥ phase. It may be understood from Eq. (3) that the critical stress of dislocation shearing into ␥ phase increases with the size (r), volume fraction (f) of ␥ -phase and antiphase boundary energy () to improve the creep resistance of the alloy. The activation energies of the alloy during creep is measured to be about 580.3 ± 20 kJ/mol, which corresponds to the better creep resistance of the alloy, and the deformation mechanisms of the alloy during creep includes the twinning, dislocations by-passing and shearing into the ␥ phase. This is well agreed with the experimental results stated above. 5. Discussion When the solution temperature increases from 1150 ◦ C to 1160 ◦ C, the average grain sizes of FGH95 alloy grow up from 10–20 m to 15–25 m, as shown in Table 3. Meanwhile, under the applied stress of 1034 MPa at 650 ◦ C, the strain rate of the alloy during steady state creep decreases from 0.0102%/h to 0.00367%/h, the creep lifetime of the alloy increases from 67 h to 104 h, as shown in Fig. 5 and Table 4. This indicates that the coarser grain size can improve the creep resistance of the alloy. The average grain size of the alloy increases as the solution temperature enhances to 1165 ◦ C, but the carbides were continuously precipitated to form the films along the boundaries as shown in Fig. 3, which decreases the cohesive strength between the grains to damage the lifetime of the alloy. Although the alloy possesses a lower strain rate of about 0.00304%/h during steady state creep, the creep lifetime of the one is only about 9 h, as shown in Fig. 5 and Table 4. After solution treated at 1160 ◦ C (dissolving temperature of ␥ phase is about 1160 ◦ C [21]), the coarser ␥ particles are completely dissolved, and the fine ␥ phase with the high volume fraction is dispersedly precipitated within the grains to eliminate the depleted zones of the fine ␥ phase, as shown in Fig. 4(a), which may hinder the dislocation movement to improve the creep resistance of the alloy. When the carbide particles are precipitated along the boundaries as shown in Fig. 7(d), the cohesive strength between the grains may be improved, due to the pinning effect of them, to restrain the boundaries slipping during creep. Therefore, when the crack is initiated and propagated along the boundaries during creep, the fracture in the alloy displays the unsmooth surface as shown in Fig. 11(c). If no carbide particles are precipitated along the boundaries, the fracture after the alloy crept up to rupture displays the surface feature of the smooth, which corresponds to the lower creep resistance of alloy due to the weaker cohesive strength between the grains [23].
6. Conclusion (1) After solution treated at 1150 ◦ C, some coarser ␥ precipitates are distributed in the wider boundary regions where appears the depleted zone of the fine ␥ -phase. After solution temperature rises to 1160 ◦ C, the coarser ␥ phase in the alloy is fully dissolved, the fine secondary ␥ phase with high volume fraction is dispersedly distributed within the grains, and the particles of (Nb, Ti)C carbide are precipitated along the boundaries. When the alloy is solution treated at 1165 ◦ C, the size of the grains is obviously grown up, and the carbides are continuously precipitated to form the films along the boundaries. (2) The carbide particles precipitated along the boundaries can effectively hinder the dislocation movement and improve the cohesive strength between the grains after the alloy is solution treated at 1160 ◦ C and twice aged, which is thought to be the main reason of the alloy possessing better creep resistance and longer lifetime. (3) The deformation mechanisms of the alloy during creep are the twinning, dislocations by-passing or shearing into the ␥ phase. The 1 1 0 super-dislocations shearing into the ␥ phase may be decomposed to form the configuration of (1/3)1 1 2 superShockleys partial dislocations and stacking fault. (4) In the later stage of creep, the deformation features of the alloy are the single or double orientations slipping of dislocations activated in the alloy. Some dislocations are piled up in the regions near the boundaries, which may bring about the stress concentration to promote the initiation and propagation of the microcracks along the boundaries, this is thought to be the fracture mechanism of the alloy during creep. References [1] V.P. Swaminathan, N.S. Cheruvu, J.M. Klein, W.M. Robinson, ASME International Gas Turbine and Aeroengine Congress, 1998, Paper No. 98-GT-510. [2] D.M. Liu, Y. Zhang, P.Y. Liu, et al., Powder Metallurgy Industry 16 (3) (2006) 1–5. [3] N.K. Park, I.S. Kim, Journal of Materials Processing Technology 111 (2) (2001) 98–102. [4] B. Flageolet, M. Jouiad, P. Villechaise, et al., Materials Science and Engineering A 399 (2005) 199–205. [5] L.W. Lherbier, W.B. Kent, The International Journal of Powder Metallurgy 26 (2) (1990) 131–137. [6] H.M. Chen, B.F. Hu, H.Y. Li, et al., The Chinese Journal of Nonferrous Metals 13 (3) (2003) 554–559. [7] S. Raujol, F. Pettinari, D. Locq, et al., Materials Science and Engineering A 387–389 (2004) 678–682. [8] S. Terzi, R. Couturier, L. Guetal, et al., Materials Science and Engineering A 483–484 (2008) 598–601. [9] B.F. Hu, H.M. Chen, K.Sh. Jin, et al., The Chinese Journal of Nonferrous Metals 14 (6) (2004) 901–906. [10] H.D. Zainul, Materials and Design 28 (2007) 1664–1667. [11] Z.Z. Lu, C.L. Liu, Z.F. Yue, Materials Science and Engineering A 395 (2005) 153–159. [12] S.G. Tian, Y. Liu, X.M. Zhao, et al., Journal of Aeronautical Materials 29 (6) (2009) 33–37. [13] L. Paul, Powder Metallurgy Superalloys (1988) 27–36.
2084
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084
[14] R.R. Unocic, G.B. Viswanathan, P.M. Sarosi, et al., Materials Science and Engineering A 483–484 (2008) 25–32. [15] L. Kovarik, R.R. Unocic, J. Li, M.J. Mills, Journal of the Minerals 61 (2) (2009) 42–48. [16] G.B. Viswanathan, P.M. Sarosi, M.F. Henry, et al., Acta Materialia 53 (2005) 3041–3057. [17] J.B. Zhou, J.X. Dong, Zh.Ch. Xu, et al., Heat Treatment of Metals 27 (6) (2002) 30–32. [18] P. Wang, J.X. Dong, L. Yang, et al., Material Review 22 (6) (2008) 61– 64.
[19] CH.CH. Jia, F.ZH. Yin, B.F. Hu, et al., Materials Science and Engineering of Powder Metallurgy 11 (3) (2006) 176–179. [20] S.G. Tian, X.F. Yu, J.H. Yang, et al., Materials Science and Engineering A 379 (2004) 141–147. [21] Y.W. Zhang, Y. Zhang, F.G. Zhang, et al., Transactions of Materials and Heat Treatment 23 (3) (2002) 72–75. [22] J.SH. Zhang, High Temperature Deformation and Fracture of Material [M], Science Press, Beijing, 2007, pp. 102–105. [23] S.G. Tian, Z.H.G. Zhao, L.Q. Chen, et al., Journal of Aeronautical Materials 30 (5) (2010) 14–18.
Contents lists available at ScienceDirect
Materials Science and Engineering A journal homepage: www.elsevier.com/locate/msea
Microstructure and creep behavior of FGH95 nickel-base superalloy Tian Sugui a,∗ , Xie Jun a , Zhou Xiaoming b , Qian Benjiang a , Lun Jianwei a , Yu Lili a , Wang Wuxiang b a b
School of Materials Science and Engineering, Shenyang University of Technology, Shenyang 110870, China Beijing Institute of Aeronautical Materials, Beijing 100095, China
a r t i c l e
i n f o
Article history: Received 30 July 2010 Received in revised form 10 November 2010 Accepted 10 November 2010
Keywords: FGH95 nickel-base superalloy Microstructure Creep Deformation mechanism Fracture feature
a b s t r a c t By means of the measurement of creep curves and the microstructure observations, the influence of the solution temperatures on the creep behavior of FGH95 nickel-base superalloy is investigated. Results show that, after solution treated at 1150 ◦ C, some coarser ␥ precipitates are distributed in wider boundary regions where no fine ␥ phase is precipitated. As the solution temperature is raised to 1165 ◦ C, the grain size of the alloy increases obviously, and the carbide is continuously precipitated to form the film along the boundaries. When the alloy is solution treated at 1160 ◦ C, the coarser ␥ phase in the alloy is fully dissolved, the fine ␥ phase with higher volume fraction is dispersedly distributed within the grains, and some particles of (Nb, Ti)C are precipitated along the grain boundaries, which can effectively hinder the grain boundary slipping and dislocation moving. Thereby, the alloy displays a better creep resistance under the applied stress of 1034 MPa at 650 ◦ C. The deformation mechanism of the alloy during creep is twinning, dislocation shearing or bypassing the ␥ phase, and the 1 1 0 super-dislocation which shears into the ␥ phase may be decomposed to form the configuration of (1/3)1 1 2 super-Shockleys partial and stacking fault. In the later stage of creep, the deformation features of the alloy are the single and double orientations slipping of dislocations activated in the alloy. As the creep goes on, some dislocations piled up in the regions near the boundaries may bring the stress concentration to promote the initiation and propagation of micro-cracks, which is thought to be the fracture mechanism of the alloy during creep. © 2010 Published by Elsevier B.V.
1. Introduction With the development of aerospace and ground transportation industry, the properties of aerospace turbines are increasingly required to be improved, and especially, the turbine disks of aeroengines are required to be of high temperature tolerance and creep resistance [1]. Traditional wrought superalloys can hardly meet the requirements of aerospace turbine disks for their poor temperature tolerance and loading capacity resulted from their serious composition segregation in ingots and poor hot processability [2–5], especially the weaker cohesive force of grain boundaries [6–8]. FGH95 alloy is a nickel-base superalloy with higher alloying degree and volume fraction of ␥ -phase [6,9]. Compared with wrought superalloys, FGH95 superalloy has the capabilities of higher temperature tolerance and loading tolerance, due to their excellent characteristics, such as uniform chemical composition and fine grain size. Therefore, the alloy is an excellent material used for preparing the advanced aerospace turbine disks with high thrust-weight ratio [10,11]. The microstructure of FGH95 superalloy consists of ␥ matrix, ␥ and carbide phases. Various size, morphology and distribu-
∗ Corresponding author. Tel.: +86 24 25494089; fax: +86 24 25496768. E-mail address: [email protected] (T. Sugui). 0921-5093/$ – see front matter © 2010 Published by Elsevier B.V. doi:10.1016/j.msea.2010.11.038
tion of ␥ phase can be obtained in the alloy by different heat treatment regimes [12,13]. The deformation mechanism of the polycrystalline Ni-base superalloys during creep includes twinning, dislocations by-passing or shearing into the ␥ phase [14–16]. Actually, the mechanical properties and creep behaviors of the alloy are related to the quantities, morphology and distribution of ␥ -phase, and especially, the configuration of the boundary and carbides have an important effect on the creep resistance of the alloy [17]. For example, after the alloy is solution treated for cooling in molten salt, the total number and size of secondary ␥ phase increase, which can effectively improve the plasticity of the alloy at high temperature [18,19]. Because the various microstructures in the alloy may be obtained by different heat treatment regimes, it is very important to understand the influence of heat treatment regimes on the microstructure and creep resistance of the alloy. Although some literatures report the creep behaviors and deformation features of the powder superalloys, the influence of heat treatment regimes on the configuration and distribution of ␥ -phase and carbide in FGH95 superalloy is still not clear. In this paper, the alloy is solution treated at different temperatures, and then the creep properties are measured and the microstructure is observed by using SEM and TEM for investigating the influences of the solution temperatures on the boundary morphologies and distribution of the ␥ and carbide phases. Addi-
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084 Table 1 Heat treatment regime of FGH95 powder superalloy. Solution treated
solution temperature on the microstructure and the creep feature of the alloy.
Cooled in molten salt
Aging
583 ◦ C for 15 min
870 ◦ C × 1 h + 650 ◦ C × 24 h
◦
3. Experimental results and analysis
1150 C for 1 h 1160 ◦ C for 1 h 1165 ◦ C for 1 h
3.1. Influence of solution temperature on the microstructure
Table 2 Chemical composition of FGH95 powder alloy (mass fraction, %). C
B
2077
Cr
0.060 0.012 12.98
Co
Al
Ti
W
Mo
Nb
Ni
8.00
3.48
2.55
3.40
3.40
3.50
Bal.
tionally, the deformation mechanism and fracture feature of the alloy during creep are briefly discussed. 2. Experimental procedure FGH95 powder particles of the nickel-base superalloy with the size about 150 mesh were put into a stainless steel can for pretreating at 1050 ◦ C for 4 h. The can containing FGH95 alloy powders was hot isostatic pressing treated for 4 h under the conditions of 1120 ◦ C and 120 MPa. The various solution temperatures are selected for investigating the influence of solution temperatures on the microstructure and creep properties of the alloy, and the cooled rate of the specimen in molten salt is measured to be about 110 ◦ C/min. The selected heat treatment regimes of the alloy are listed in Table 1, and the error ranges of the used heating furnace are ±2 ◦ C. The chemical composition of FGH95 alloy is shown in Table 2. By means of the anode selective dissolving method, the volume fraction of ␥ -phase in FGH95 alloy is measured to be about 47%. After heat treated at different conditions, the ingot of FGH95 alloy was cut into the specimens with the cross-section of 4.5 mm × 2.5 mm and the gauge length of 20 mm. Uniaxial constant load tensile testing was performed, in a GWT504-model creep testing machine, for measuring creep curves under the experimental conditions of 1020–1050 MPa and 650–670 ◦ C. The yield strength of FGH95 alloy is measured to be 1110 MPa at 650 ◦ C. The strain data of the alloy at different conditions were measured with an extensometer for portraying the creep curves, twice of each creep testing were conducted to ensure the statistical confidence. The microstructures of FGH95 alloy at different states were observed under SEM and TEM for investigating the influence of the
After the alloy was solution treated at 1150 ◦ C and twice aged, the microstructure of the alloy consists of the ␥ and ␥ phases, and the average grain size of the alloy is about 10–20 m, as shown in Fig. 1(a), indicating that some coarser ␥ -particles are precipitated in the wider boundary regions, and the average size of the coarser ␥ phase is about 1–2.5 m. The magnified morphology of the alloy is shown in Fig. 1(b), indicating that significant amount of the fine ␥ particles is dispersedly distributed within the grains, the size of the ones was about 0.1–0.3 m. And no fine ␥ -particles were precipitated in the regions near the coarser ␥ -phase, the regions were defined as the depleted zone of the fine ␥ -phase as marked by the arrow in Fig. 1(b), the magnified morphology of the depleted zone of ␥ -phase is marked by the arrow in Fig. 1(c). As the solution temperature raises to 1160 ◦ C, the grain boundaries appear obviously, and the average size of the grains in the alloy is about 15–25 m as shown in Fig. 2(a). Compared to Fig. 1(a), the coarser ␥ -precipitates in the boundary regions disappear, and the white particles with size about 0.2 m are precipitated within the grains and along the boundaries as marked by the arrow in Fig. 2(a). The magnified morphology of the alloy is shown in Fig. 2(b), indicating that the coarser ␥ particles and depleted zone of the fine ␥ phase have disappeared, and the secondary ␥ phase is dispersedly distributed within the grains. The grain boundary in the alloy is marked by longer arrow and some particles are homogeneously precipitated along the boundaries and within the grains as marked by the short arrows in Fig. 2(b). As the solution temperature increases to 1165 ◦ C, the grain sizes in the alloy increase, the linear-like boundaries appear in the alloy, and the films of the white phase are continuously precipitated along the boundaries as marked by the arrows in Fig. 3(a). The magnified morphology of the white phase is shown in Fig. 3(b), indicating that white particles are continuously precipitated to form the films along the boundaries as marked by the arrow in Fig. 3(b), and significant amount of the fine secondary ␥ phase is precipitated within the grains, no depleted zones of the fine ␥ phase are detected in the alloy. The grain sizes after the alloy was solution treated at various temperatures are measured as listed in Table 3. Moreover, by means of composition analysis under SEM/EDS, it is indicated that
Fig. 1. Microstructure after the alloy solution treated at 1150 ◦ C. (a) After solution treated at 1150 ◦ C, some primary ␥ phase precipitated in the wider boundary regions, (b) fine ␥ phase distributed dispersedly within the grains, and the depleted zones of the fine ␥ phase marked by arrow, and (c) the magnified morphology of the depleted zone of ␥ -phase.
2078
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084
Fig. 2. Microstructure after the alloy solution treated at 1160 ◦ C. (a) After solution treated at 1160 ◦ C, the coarser ␥ phase disappearing and the grain boundaries appearing obviously in the alloy and (b) the secondary ␥ phase distributed dispersedly within the grain, and some particles precipitated in the alloy as marked by the arrows.
Fig. 3. Microstructure after the alloy solution treated at 1165 ◦ C. (a) After solution treated at 1165 ◦ C, the average grain size of alloy increasing obviously, and films of white phase precipitated along the boundaries as marked by arrows and (b) the secondary ␥ phase distributed dispersedly within the grains, and films of the white phase distributed along the grain boundaries. Table 3 Grain sizes of the alloy solution treated at different temperatures. Solution temp. (◦ C) Average grain sizes (m)
1150 10–20
1160 15–25
1165 20–40
the elements Nb, Ti and C are richer included in the white particles which are located within the grain and boundary regions as shown in Figs. 2 and 3, respectively. After the alloy was solution treated at 1160 ◦ C, the fine ␥ particles with the size of about 0.1 m are dispersedly precipitated within the grains as shown in Fig. 4(a), the particles can effectively hinder the dislocation movement to enhance the creep resistance of the alloy. The smaller space between the fine ␥ -particle is mea-
sured to be about 0.03 m, the bigger space between the fine ␥ -particles is measured to be about 0.12 m as marked by letters L1 and L2 in Fig. 4(a), respectively. It is indicated by TEM/EDS analysis that the elements Nb, Ti and C are richer included in the carbide particle which are located in the boundary as shown in Fig. 4(b), and the particle is identified as (Nb, Ti)C phase by means of the diffraction spots analysis as marked in Fig. 4(c). 3.2. Influence of solution temperature on creep behaviors Under the applied stress of 1034 MPa at 650 ◦ C, creep curves of the alloy after solution treated at different temperatures are shown in Fig. 5. The creep curve of the alloy after solution treated at 1150 ◦ C
Fig. 4. Morphologies of ␥ and carbide phases. (a) Fine ␥ phase precipitated dispersedly within the grain, (b) carbide particles precipitated along the boundaries, and (c) SAD patterns.
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084
where ε˙ ss is the strain rate during steady state creep, A is a constant related to the microstructure, A is the applied stress, n is the apparent stress exponent, R is the gas constant, T is the absolute temperature, and Qa is the apparent creep activation energy. After solution treated at various temperatures, the creep curves of the alloy in the ranges of 650–670 ◦ C and 1020–1050 MPa were measured. The dependence of the strain rates of the alloy during steady-state creep on the applied temperatures and stresses is shown in Fig. 6. Fig. 6(a) shows the relationship between the strain rates and the temperatures under the applied stress of 1034 MPa, and the dependence of the strain rates on the applied stresses at 650 ◦ C is shown in Fig. 6(b). According to the data during the steady state creep, the creep activation energies and stress exponents of FGH95 alloy, which is solution treated at 1150 ◦ C and 1160 ◦ C, are calculated to be QA = 510.1 ± 20 kJ/mol, QB = 580.3 ± 20 kJ/mol and nA = 15.4, nB = 14.1, respectively. After solution treated at different temperatures, the creep lifetimes of the alloy under the applied various stresses and temperatures are measured as listed in Table 4. It may be understood from Table 4 that, under the applied stress of 1034 MPa at 650 ◦ C, the lifetime of the alloy solution treated at 1150 ◦ C is measured to be 67 h. As the solution temperature enhances to 1160 ◦ C, the lifetime of the alloy increases to 104 h. When the solution temperature raises further to 1165 ◦ C, the lifetime of the alloy decreases rapidly to 9 h, as shown in Table 4. This indicates that the solution temperature has an obvious influence on the creep lifetimes of the alloy. After solution treated at 1150 ◦ C, the lifetime of the alloy is about 95 h under the applied stress of 1020 MPa at 650 ◦ C, and the strain rate during steady state creep is measured to be about 0.00827%/h. The lifetime of the alloy decrease to 67 h as the applied stress increase to 1034 MPa at 650 ◦ C, and the creep lifetime of the alloy decreases to 37 h as the applied stress increases to 1050 MPa at 650 ◦ C. And the alloy solution treated at 1160 ◦ C displays also the similar regularity under the experimental conditions. This indicates that the alloy has an obvious sensitivity on the applied stresses and temperatures.
4.0
Strain, ε (%)
o
A - solution at 1150 C o 3.2 B - solution at 1160 C o C - solution at 1165 C o 2.4 T -- 650 C σ -- 1034 MPa 1.6 0.8
A
C
0.0 0
20
40
B
60
80
100
Time, (h) Fig. 5. Under the applied stress of 1034 MPa at 650 ◦ C, creep curves of the alloy solution treated at different temperatures.
is marked by letter A in Fig. 5, through which the strain rate of the alloy during steady state creep is measured to be 0.0102%/h, the lasting time is about 40 h, and the creep lifetime of the alloy is measured to be 67 h. The creep curve after the alloy solution treated at 1160 ◦ C is marked by letter B in Fig. 5, indicating that the alloy displays a lower strain rate during steady-state creep, and the creep lifetime of the alloy is measured to be 104 h. When the solution temperature increases to 1165 ◦ C, the creep lifetime of the alloy is measured to be about 9 h, as marked by letter C in Fig. 5. Transient strain of the alloy occurs when the loading is applied at high temperature, and the strain rate of the alloy decreases as the creep goes on. The strain rate keeps constant once the creep enters the steady state stage, therefore, the strain rate of the alloy during steady state creep may be expressed by Dorn’s law as follows:
Q a
ε˙ ss = AAn exp −
(1)
RT -3.5
a
σ − 1034 ΜPa
-4.0
-4.0
QA = 510.1+ 20 kJ/mol ln( εss)
ln( εss)
-5.5
QB = 580.3 + 20 kJ/mol
-6.0
1.074
1.080
1.086
o
T -- 650 C nA = 15.4
-5.0 -5.5
nB = 14.1
-6.0
-6.5 1.068
b
-4.5
-4.5 -5.0
2079
1.092
-6.5
1.098
1/T( 10-3K-1)
6.928
6.936
6.944
ln ( σ)
6.952
6.960
Fig. 6. Dependence of the strain rate during steady state creep on the applied temperatures and stresses for the alloy solution treated by different temperatures. (a) Strain rate vs. temperatures at 1034 MPa and (b) strain rate vs. the applied stress at 650 ◦ C.
Table 4 Effect of solution temperature on stress rupture properties of FGH95 alloy. Solution temp. (◦ C)
650 ◦ C
1034 MPa
1020 MPa
1150 1160 1165
1034 MPa
660 ◦ C
1050 MPa
670 ◦ C
tf (h)
ı (%)
ε˙ (%/h)
tf (h)
ı (%)
ε˙ (%/h)
tf (h)
ı (%)
ε˙ (%/h)
tf (h)
ı (%)
ε˙ (%/h)
tf (h)
ı (%)
ε˙ (%/h)
95 156 –
4.2 3.7 –
0.00827 0.00307 –
67 104 9
3.4 2.8 1.3
0.0102 0.00367 0.00304
37 51 –
3.2 2.1 –
0.0129 0.00456 –
26 40 –
3.6 2.2 –
0.0208 0.00825 –
12 17 –
2.3 1.7 –
0.0418 0.0179 –
2080
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084
Fig. 7. Microstructure after the alloy (solution treated at 1160 ◦ C) crept for 104 h up to fracture at 650 ◦ C/1034 MPa. (a) Dislocation loops in the alloy, (b) carbide particles precipitated uniformly within the grain, (c) morphology of the super-dislocation shearing into ␥ phase marked by black arrow, the stacking fault in the region A and dislocation tangles piled up in the B region near the boundary, (d) carbide particles precipitated along the boundary, and dislocation tangles piled up in the regions near the boundaries.
3.3. Deformation features of alloy during creep After solution treated at 1160 ◦ C, the morphology of the alloy crept for 104 h up to rupture under the applied stress of 1034 MPa at 650 ◦ C is shown in Fig. 7. Fig. 7(a) displays the feature of the stacking fault as marked by white arrow, and two (1/3)1 1 2 superShockleys partials are located on the two sides of the stacking fault [15]. Some dislocation loops with various sizes appear clearly in the ␥ matrix of the alloy, the smaller dislocation loop is marked with black arrow in Fig. 7(a), and larger dislocation loop is located in the region above the stacking fault. It can be deduced by analysis that the 1 1 0 super-dislocation which shears into the ␥ or ␥ -phase may be decomposed to form the configuration of two (1/3)1 1 2 super-Shockleys partials and the stacking fault [20]. Some blocky carbides are precipitated in local areas of the alloy as marked by black arrow in Fig. 7(b). The grain boundary in the alloy is marked by white arrow in Fig. 7(c), and dislocation tangles are piled up in the region B near the boundary, which suggests that the boundaries may effectively hinder the dislocation movement during creep. Besides, the morphology of 1 1 0 super-dislocation shearing into the ␥ -phase is marked by the black arrow, and the stacking fault formed from the dislocation decomposition is marked by the letter A in Fig. 7(c). The blocky carbides are precipitated along the boundary as marked by the black arrow in Fig. 7(d),
and some dislocations are piled up in the regions near the carbides, which indicates that the carbide particles can effectively hinder the dislocation movement. But, the fact that significant amount of dislocations is piled up in the regions near the boundaries can cause the stress concentration to promote the initiation and propagation of the microcrack along the boundaries as the creep goes on. After the alloy was crept up to fracture under the applied stress of 1034 MPa at 650 ◦ C, the feature of the twinning deformation is shown in Fig. 8(a), and the twinning plane is identified as (1 1 1) plane by means of the diffraction pattern analysis as shown in Fig. 8(b). The blocky carbide is precipitated along the grain boundary as marked by shorter arrow in Fig. 9, and some micro-twinning is detected in the alloy as marked by black arrow. In another local area, the morphology and diffraction spots of the micro-twinning are shown in Fig. 9(b) and (c), this is well agreement with the result in literature [14]. The one end of the micro-twinning is stopped at the grain boundary, which suggests that the grain boundaries have an obvious effect on hindering the twinning deformation. 3.4. Fracture feature of alloy during creep After solution treated at 1160 ◦ C, the morphology of the alloy crept for different times under the applied stress of 1034 MPa at 650 ◦ C is shown in Fig. 10.
Fig. 8. Twinning morphology in FGH95 alloy during creep at 650 ◦ C/1034 MPa. (a) Twinning and (b) SAD patterns.
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084
2081
Fig. 9. Morphologies of the carbide precipitated along the boundary and micro-twinning formed in the alloy during creep under the applied stress of 1034 MPa at 650 ◦ C. (a) Carbide precipitated along the boundary, (b) micro-twinning, and (c) SAD patterns.
Fig. 10. Morphologies of slipping traces on the surface of the specimen (solution treated at 1160 ◦ C) crept for different times. (a) After crept for 60 h, a few slipping traces appearing within the different grains and (b) crept for 80 h, significant amount of the slipping traces appearing on the surface of the sample.
After the alloy is crept for 60 h, the morphology of the slipping traces on the sample surface is shown in Fig. 10(a), which displays the feature of the single orientation slipping within the grain. But the slipping traces with different directions appear within different grains, and the intersected slipping traces appear in the boundary region as marked by the arrow in Fig. 10(a), which indicates that the boundary may change the direction of the slipping traces. When crept for 80 h, the amount of the slipping trace on the sample surfaces increases obviously, as shown in Fig. 10(b), and some white particles are uniformly precipitated within the grains.
Significant amount of dislocations is piled up in the regions near the boundary and carbides during creep, as shown in Fig. 7(d), which suggests that the carbides and boundaries can effectively hinder the dislocation movement. The dislocations piled up in the regions near the boundaries may bring about the stress concentration, as the creep goes on, to promote the initiation of the micro-cracks along the boundary when the value of the stress concentration exceeds the yield strength of the boundary, as marked by the arrow in Fig. 11(a). The number of the slipping traces in the sample surface increases as the creep goes on, as shown in Fig. 11(b), and the
Fig. 11. Morphologies of the slipping traces on the surface of alloy (solution treated at 1160 ◦ C) crept up to fracture. (a) Micro-crack initiated along the boundary as marked by the arrow, (b) slipping traces in the surface, and (c) feature of cracks propagated along the boundaries.
2082
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084
Fig. 12. After solution treated at 1165 ◦ C, morphology of the alloy crept for 9 h up to fracture. (a) Crack initiated along the boundary and (b) carbide film reserved between the cracks.
Fig. 13. Schematic diagram of dislocation bowing out in the congregated regions of the ␥ particles.
slipping traces with the double-orientation feature appear within the grains, the slipping steps are displayed in the region near the boundaries as marked with the white arrow, and some cracks appear in the region near the boundary as marked with the shorter arrow in Fig. 11(b). Furthermore, when significant amount of dislocations move to the region near the boundaries again to produce the stress concentration, this may promote the propagation of the crack along the boundary, as shown in Fig. 11(c), which indicates that the fracture originated from the cracks initiating and propagating along the boundaries displays a non-smooth surface due to the pinning effect of the carbide particles precipitated along the boundaries. Therefore, it may be deduced that the carbide particles precipitated along the boundaries may restrain the cracks propagation along the boundaries for enhancing the creep resistance of the alloy. After solution treated at 1165 ◦ C and twice aged, the surface morphology of the alloy crept for 9 h up to rupture under the applied stress of 1034 MPa at 650 ◦ C is shown in Fig. 12, which illustrates that only a few slipping trace appears on the surface of the sample, and some micro-cracks are initiated along the boundaries vertical to the applied stress axis as marked by arrow in Fig. 12(a). After the sample is polished and eroded, the morphology of the crack propagated along the boundaries is shown in Fig. 12(b). It may be understood from Fig. 12(b) that the fracture after the crack is propagated displays the smooth surface, and the carbide film is reserved within the crack between the grains as marked by arrow in Fig. 12(b), which suggests that once the carbide is continuously precipitated to form the film along the boundary, which may weaken the cohesive strength between the grains to damage the creep lifetimes of the alloy. Moreover, it is identified by means of composition analysis under SEM/EDS that the elements Nb, Ti, C and O are richer included in the white particles on the surface of the samples, as shown in Figs. 10, 11 and 12, respectively, therefore, it is thought that the white particles on the surface of the samples are the oxides of the elements Nb, Ti and C.
4. Analysis on deformation features of the alloy during creep Significant amount of the fine ␥ particles is precipitated within the grains, which may effectively hinder the dislocation movement. When the deformed dislocations move over the ␥ phase during creep, the dislocation loops are kept around the ␥ particles as shown in Fig. 7(a), which suggests that the deformation feature of the alloy during creep is the dislocations moving over the ␥ -phase by Orowan bypasses mechanism. It is reasonable consideration that the various spaces between the ␥ particles appear in different regions, the dislocations may bow out along the wider channels between the two ␥ particles during creep, and the bowing dislocations move over the ␥ -phase by Orowan mechanism to encounter for forming the dislocation loops, as marked with black arrow in Fig. 7(a). When the dislocations bow out along the channels during creep, the applied stress which is enough to overcome the Orowan resistance can be expressed as follows: or =
·b L
(2)
where is the shear modulus, b is the Burgers vector, and L is the space between two ␥ particles. This indicates that the resistance of the dislocation movement increases with the diminishing of the space between the ␥ -particles, and the resistance of alloy enhances with the volume fraction of ␥ -particles. The space between the ␥ particles diminishes when some ␥ particles are congregated together, the smaller space is measured to be L1 = 0.03 m as shown in Fig. 4(a), but the bigger space between the ␥ -particles is measured to be L2 = 0.12 m. Compared to the bigger space, the dislocations bowing out in the regions with smaller space needs fourfold Orowan resistance according to the formula (2), therefore, it is difficult for the dislocations to bow out along the channels with smaller space, but they can bow out along the channels with larger space to form the larger loops as shown in
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084
Fig. 7(a). When the dislocations bow out along the channels with bigger space (L), the formation process of the bigger dislocation loops is schematically shown in Fig. 13. In the later stage of creep, significant amount of dislocations move to the region near the ␥ particles to generate stress concentration, and when the stress value originated from the stress concentration exceeds the yield strength of ␥ phase, the 1 1 0 super-dislocation may shear into the ␥ -phase, as marked by black arrow in Fig. 7(c). Furthermore, the 1 1 0 super-dislocations may be decomposed to form the configuration of two (1/3)1 1 2 superShockleys partials and the stacking fault, as marked by letter A in Fig. 7(c). The critical stress of dislocation shearing into ␥ phase increases with the yield strength. And the critical stress () can be expressed as follows [22]: cs =
APB b
0.3 ·
APB
T
·f ·r
2083
Because the boundaries and the carbide particles can effectively block the dislocation movement, and especially, the carbides may restrain the boundaries slipping during creep, it may be concluded that the carbide particles precipitated along the boundaries have an important effect on improving the creep resistance of the alloy. Although the carbide particles precipitated along the boundaries can improve the cohesive strength of the boundaries, the microcracks are still initiated and propagated along the boundaries, which suggests that the boundaries are still the weaker regions for causing fracture of the alloy during creep. And once, the carbide is continuously precipitated to form the film along the boundary, which may weaken the cohesive strength between the grains to damage the creep lifetimes of the alloy. The analysis is in agreement with the experimental results stated above.
1/2 (3)
where T is the dislocation line tension, r is the radius of the ␥ particle, b is the Burgers vector, f is the volume fraction of ␥ phase, and APB is the antiphase boundary energy originated from the dislocation shearing ␥ phase. It may be understood from Eq. (3) that the critical stress of dislocation shearing into ␥ phase increases with the size (r), volume fraction (f) of ␥ -phase and antiphase boundary energy () to improve the creep resistance of the alloy. The activation energies of the alloy during creep is measured to be about 580.3 ± 20 kJ/mol, which corresponds to the better creep resistance of the alloy, and the deformation mechanisms of the alloy during creep includes the twinning, dislocations by-passing and shearing into the ␥ phase. This is well agreed with the experimental results stated above. 5. Discussion When the solution temperature increases from 1150 ◦ C to 1160 ◦ C, the average grain sizes of FGH95 alloy grow up from 10–20 m to 15–25 m, as shown in Table 3. Meanwhile, under the applied stress of 1034 MPa at 650 ◦ C, the strain rate of the alloy during steady state creep decreases from 0.0102%/h to 0.00367%/h, the creep lifetime of the alloy increases from 67 h to 104 h, as shown in Fig. 5 and Table 4. This indicates that the coarser grain size can improve the creep resistance of the alloy. The average grain size of the alloy increases as the solution temperature enhances to 1165 ◦ C, but the carbides were continuously precipitated to form the films along the boundaries as shown in Fig. 3, which decreases the cohesive strength between the grains to damage the lifetime of the alloy. Although the alloy possesses a lower strain rate of about 0.00304%/h during steady state creep, the creep lifetime of the one is only about 9 h, as shown in Fig. 5 and Table 4. After solution treated at 1160 ◦ C (dissolving temperature of ␥ phase is about 1160 ◦ C [21]), the coarser ␥ particles are completely dissolved, and the fine ␥ phase with the high volume fraction is dispersedly precipitated within the grains to eliminate the depleted zones of the fine ␥ phase, as shown in Fig. 4(a), which may hinder the dislocation movement to improve the creep resistance of the alloy. When the carbide particles are precipitated along the boundaries as shown in Fig. 7(d), the cohesive strength between the grains may be improved, due to the pinning effect of them, to restrain the boundaries slipping during creep. Therefore, when the crack is initiated and propagated along the boundaries during creep, the fracture in the alloy displays the unsmooth surface as shown in Fig. 11(c). If no carbide particles are precipitated along the boundaries, the fracture after the alloy crept up to rupture displays the surface feature of the smooth, which corresponds to the lower creep resistance of alloy due to the weaker cohesive strength between the grains [23].
6. Conclusion (1) After solution treated at 1150 ◦ C, some coarser ␥ precipitates are distributed in the wider boundary regions where appears the depleted zone of the fine ␥ -phase. After solution temperature rises to 1160 ◦ C, the coarser ␥ phase in the alloy is fully dissolved, the fine secondary ␥ phase with high volume fraction is dispersedly distributed within the grains, and the particles of (Nb, Ti)C carbide are precipitated along the boundaries. When the alloy is solution treated at 1165 ◦ C, the size of the grains is obviously grown up, and the carbides are continuously precipitated to form the films along the boundaries. (2) The carbide particles precipitated along the boundaries can effectively hinder the dislocation movement and improve the cohesive strength between the grains after the alloy is solution treated at 1160 ◦ C and twice aged, which is thought to be the main reason of the alloy possessing better creep resistance and longer lifetime. (3) The deformation mechanisms of the alloy during creep are the twinning, dislocations by-passing or shearing into the ␥ phase. The 1 1 0 super-dislocations shearing into the ␥ phase may be decomposed to form the configuration of (1/3)1 1 2 superShockleys partial dislocations and stacking fault. (4) In the later stage of creep, the deformation features of the alloy are the single or double orientations slipping of dislocations activated in the alloy. Some dislocations are piled up in the regions near the boundaries, which may bring about the stress concentration to promote the initiation and propagation of the microcracks along the boundaries, this is thought to be the fracture mechanism of the alloy during creep. References [1] V.P. Swaminathan, N.S. Cheruvu, J.M. Klein, W.M. Robinson, ASME International Gas Turbine and Aeroengine Congress, 1998, Paper No. 98-GT-510. [2] D.M. Liu, Y. Zhang, P.Y. Liu, et al., Powder Metallurgy Industry 16 (3) (2006) 1–5. [3] N.K. Park, I.S. Kim, Journal of Materials Processing Technology 111 (2) (2001) 98–102. [4] B. Flageolet, M. Jouiad, P. Villechaise, et al., Materials Science and Engineering A 399 (2005) 199–205. [5] L.W. Lherbier, W.B. Kent, The International Journal of Powder Metallurgy 26 (2) (1990) 131–137. [6] H.M. Chen, B.F. Hu, H.Y. Li, et al., The Chinese Journal of Nonferrous Metals 13 (3) (2003) 554–559. [7] S. Raujol, F. Pettinari, D. Locq, et al., Materials Science and Engineering A 387–389 (2004) 678–682. [8] S. Terzi, R. Couturier, L. Guetal, et al., Materials Science and Engineering A 483–484 (2008) 598–601. [9] B.F. Hu, H.M. Chen, K.Sh. Jin, et al., The Chinese Journal of Nonferrous Metals 14 (6) (2004) 901–906. [10] H.D. Zainul, Materials and Design 28 (2007) 1664–1667. [11] Z.Z. Lu, C.L. Liu, Z.F. Yue, Materials Science and Engineering A 395 (2005) 153–159. [12] S.G. Tian, Y. Liu, X.M. Zhao, et al., Journal of Aeronautical Materials 29 (6) (2009) 33–37. [13] L. Paul, Powder Metallurgy Superalloys (1988) 27–36.
2084
T. Sugui et al. / Materials Science and Engineering A 528 (2011) 2076–2084
[14] R.R. Unocic, G.B. Viswanathan, P.M. Sarosi, et al., Materials Science and Engineering A 483–484 (2008) 25–32. [15] L. Kovarik, R.R. Unocic, J. Li, M.J. Mills, Journal of the Minerals 61 (2) (2009) 42–48. [16] G.B. Viswanathan, P.M. Sarosi, M.F. Henry, et al., Acta Materialia 53 (2005) 3041–3057. [17] J.B. Zhou, J.X. Dong, Zh.Ch. Xu, et al., Heat Treatment of Metals 27 (6) (2002) 30–32. [18] P. Wang, J.X. Dong, L. Yang, et al., Material Review 22 (6) (2008) 61– 64.
[19] CH.CH. Jia, F.ZH. Yin, B.F. Hu, et al., Materials Science and Engineering of Powder Metallurgy 11 (3) (2006) 176–179. [20] S.G. Tian, X.F. Yu, J.H. Yang, et al., Materials Science and Engineering A 379 (2004) 141–147. [21] Y.W. Zhang, Y. Zhang, F.G. Zhang, et al., Transactions of Materials and Heat Treatment 23 (3) (2002) 72–75. [22] J.SH. Zhang, High Temperature Deformation and Fracture of Material [M], Science Press, Beijing, 2007, pp. 102–105. [23] S.G. Tian, Z.H.G. Zhao, L.Q. Chen, et al., Journal of Aeronautical Materials 30 (5) (2010) 14–18.